Abstract
Ocean related activities are often supported by offshore equipment with particular power demands. These are usually deployed at remote locations and have limited space, thus small energy harvesting technologies, such as photovoltaic panels or wind turbines, are used to power their instruments. However, the inherent energy sources are intermittent and have lower density and predictability than an alternative source: wave energy. Here, we propose and critically assess triboelectric nanogenerators (TENGs) as a promising technology for integration into wave buoys. Three TENGs based on rolling-spheres were developed and their performance compared in both a "dry" bench testing system under rotating motions, and in a large-scale wave basin under realistic sea-states installed within a scaled navigation buoy. Both experiments show that the electrical outputs of these TENGs increase with decreasing wave periods and increasing wave amplitudes. However, the wave basin tests clearly demonstrated a significant dependency of the electrical outputs on the pitch degree of freedom and the need to take into account the full dynamics of the buoy, and not only that of TENGs, when subjected to the excitations of waves. This work opens new horizons and strategies to apply TENGs in marine applications, considering realistic hydrodynamic behaviors of floating bodies.

Abstract
In this paper, we address a subtopic of the broad domain of audio enhancement, namely musical audio bandwidth extension. We formulate the bandwidth extension problem using deep neural networks, where a band-limited signal is provided as input to the network, with the goal of reconstructing a full-bandwidth output. Our main contribution centers on the impact of the choice of low-pass filter when training and subsequently testing the network. For two different state-of-the-art deep architectures, ResNet and U-Net, we demonstrate that when the training and testing filters are matched, improvements in signal-to-noise ratio (SNR) of up to 7 dB can be obtained. However, when these filters differ, the improvement falls considerably and under some training conditions results in a lower SNR than the band-limited input. To circumvent this apparent overfitting to filter shape, we propose a data augmentation strategy which utilizes multiple low-pass filters during training and leads to improved generalization to unseen filtering conditions at test time.

Abstract
[No abstract available]

Abstract
The recent Covid-19 pandemic and the fact that wearing masks in public is now mandatory in several countries, created challenges in the use of face recognition systems (FRS). In this work, we address the challenge of masked face recognition (MFR) and focus on evaluating the verification performance in FRS when verifying masked vs unmasked faces compared to verifying only unmasked faces. We propose a methodology that combines the traditional triplet loss and the mean squared error (MSE) intending to improve the robustness of an MFR system in the masked-unmasked comparison mode. The results obtained by our proposed method show improvements in a detailed step-wise ablation study. The conducted study showed significant performance gains induced by our proposed training paradigm and modified triplet loss on two evaluation databases.

Abstract

Abstract